Benchmark result
qwen/qwen3-8b on Apple M5 — 24.2 tok/s
Measured with LM Studio on August 27, 2026.
What the model built
Open full screen →The coding scenario asks for a playable game in a single HTML file. This is exactly what the model returned, unedited.
How this run compares
qwen/qwen3-8b on other hardware
Reproduce this run
toolLM Studio
modelqwen/qwen3-8b
context8,192
clientv0.4.44+97
Set those in LM Studio, then:
llm-benchmark benchmark --model "qwen/qwen3-8b" --tool "LM Studio"The client prompts for context length and thinking mode, and for the KV cache dtype on Unsloth Studio. Sampling is left at the model default — the values above are what the tool reported using, not overrides the client sent.